Yardstick Research tear-sheet / AI sales cohort

Methodology · how we score · rubric weights in plain sight · vendors received this sheet seven days before publication and could flag factual errors, never rankings

Nabla

Identity

Founded: 2018 by Alexandre Lebrun, Delphine Groll, and Martin Raison - the Wit.ai team that sold their prior NLP startup to Facebook in 2015. [VENDOR-CLAIMED + THIRD-PARTY corroboration - https://www.nabla.com/team, https://siliconangle.com/2025/06/17/clinical-ai-startup-nabla-focus-agentic-automation-raising-70m-funding/]
HQ: Paris, France (22 rue Chapon) and Brooklyn, NY (1 Dock 72 Way). Dual operating presence, founder team rooted in Paris with US commercial team in New York. [VENDOR-CLAIMED + THIRD-PARTY - https://www.nabla.com/, https://tracxn.com/d/companies/nabla/__9HH3-1wWpzD4sdQ2lByXePi71KyKD3MkB_6kBcu2pQw, https://www.zoominfo.com/c/nabla/397709641]
Legal entity: Nabla Technologies, Inc. listed by aggregators for the US entity; original French operating entity behind the Paris HQ referred to as Nabla on press releases. Single-vs-sister-entity structure worth right-of-reply confirmation.
Funding: $120M total raised across Seed through Series C. Most recent: $70M Series C closed June 17, 2025, led by HV Capital (Germany), with Highland Europe and DST Global as new investors, plus existing backers Cathay Innovation and Tony Fadell's Build Collective. Valuation not disclosed publicly (Series B in January 2024 was at ~$180M post-money). 5x revenue growth in the 6 months prior to Series C close. [VENDOR-CLAIMED + THIRD-PARTY corroboration - https://www.nabla.com/press-release/nabla-raises-70m-series-c-to-deliver-agentic-ai-to-the-heart-of-clinical-workflows-bringing-total-funding-to-120m, https://www.statnews.com/2025/06/17/nabla-raises-70-million-ambient-market-heats-up/, https://news.crunchbase.com/ai/nabla-ai-copilot-agents-healthcare-funding/]
Headcount: ~143-144 as of late February 2026; doubled headcount across the Series C window. Distributed across Paris HQ, Brooklyn NY office, and remote contributors across 6 continents. [THIRD-PARTY - https://tracxn.com/d/companies/nabla/__9HH3-1wWpzD4sdQ2lByXePi71KyKD3MkB_6kBcu2pQw]
Cap-table notables: HV Capital, Highland Europe, DST Global, Cathay Innovation, Build Collective (Tony Fadell), ZEBOX Ventures. Tony Fadell (Nest / iPod) cap-table presence + Yann LeCun (Meta Chief AI Scientist) on the advisor + AMI Labs partnership roster give Nabla unusual consumer-product and frontier-AI gravity vs the US-VC-heavy cohort peers.
Leadership:
Alexandre Lebrun - Co-founder & CEO (ex-Wit.ai / Facebook AI Research)
Delphine Groll - Co-founder & COO
Martin Raison - Co-founder & CTO (ex-Wit.ai / Facebook AI Research)
Laurent Landowski - Chief Product Officer
Ed Lee, MD, MPH - Chief Medical Officer
Matthew Sakumoto, MD - Chief Clinical Product Officer (appointed February 10, 2026)
Chloe Thomas - CFO
Priscilla Pennington - VP Sales
Christina Boulier - VP Customer Success
Amanda Wolfe - VP Marketing
Advisors: Yann LeCun (Meta), Megan Mahoney MD (Stanford), Haipeng Zhang DO
[VENDOR-CLAIMED - https://www.nabla.com/team]
Recent news (last 12 months):
May 14, 2026 - UC San Diego Health Becker's webinar deployment cited. [VENDOR-CLAIMED - https://www.nabla.com/press]
April 21, 2026 - LCMC Health follow-on customer-success case study published. [VENDOR-CLAIMED]
February 10, 2026 - Dr. Matt Sakumoto appointed Chief Clinical Product Officer. [VENDOR-CLAIMED]
February 5, 2026 - M Health Fairview selects Nabla's combined ambient AI + dictation solution. [VENDOR-CLAIMED]
January 15, 2026 - Aultman Health System scales Nabla through Oracle Cerner integration - first named-customer Cerner-deep production deployment. [VENDOR-CLAIMED]
December 18, 2025 - Exclusive partnership announced with Yann LeCun's Advanced Machine Intelligence (AMI Labs) for world-model architecture in autonomous clinical agents. [VENDOR-CLAIMED]
December 3, 2025 - LCMC Health system-wide ambient AI selection. [VENDOR-CLAIMED]
November 26, 2025 - NEJM AI publishes "Ambient AI Scribes in Clinical Practice: A Randomized Trial" - the UCLA 238-physician three-arm pragmatic RCT. Nabla shows statistically significant 9.5% reduction in time-in-note vs control (95% CI: -17.2% to -1.8%, P=0.02); 7.8% faster than Microsoft DAX Copilot (P=0.05). Only Nabla-specific peer-reviewed RCT in the cohort with head-to-head comparator. [THIRD-PARTY - https://ai.nejm.org/doi/abs/10.1056/AIoa2501000, https://pmc.ncbi.nlm.nih.gov/articles/PMC12768499/]
October 9, 2025 - 30+ provider groups adopt Nabla via athenahealth Marketplace in first year. [VENDOR-CLAIMED]
October 2025 - Nabla Connect launched: plug-and-play OEM embedding API; powers NextGen Ambient Assist + Greenway Health ambient documentation as white-label. [VENDOR-CLAIMED]
June 17, 2025 - $70M Series C close at $120M cumulative led by HV Capital. [VENDOR-CLAIMED]
Latent risk: Historic 2020 research artifact persists on Google. Prior to pivoting fully into ambient documentation, the Nabla team published research showing GPT-3 in a medical-Q&A configuration responded "I think you should" to a mock patient who asked "Should I kill myself?" The Nabla team's own published conclusion was that GPT-3 was unsuitable for medical use and they did not ship the product. The headline persists in search results and may be surfaced by procurement teams unfamiliar with the context (research experiment, not production deployment, no patient harm, published the failure, pivoted). Brand-image risk rather than product risk. [THIRD-PARTY - https://www.theregister.com/2020/10/28/gpt3_medical_chatbot_experiment/]
Archetype: Mid-market-to-enterprise ambient AI clinical documentation platform with two structural differentiators - EHR-agnostic + browser-based architecture (only cohort vendor that ships a productized OEM embedding API; powers NextGen Ambient Assist + Greenway Health as white-label), and zero-storage privacy posture (audio + transcripts retained only briefly in clinician browser; 14-day note default retention; no audio storage by default). Sits between the enterprise Epic-deep tier (Abridge / DAX Copilot / Ambience) and the SMB self-serve tier (Freed / Heidi). Competing on EHR breadth, the only Nabla-specific peer-reviewed RCT in the cohort, 35+ languages (broadest in cohort), and a privacy posture that is rare even among healthcare-grade vendors.

Total score: 78.75 / 100

Stage fit:
Foundation (<40 readiness): yes - More so than any cohort peer. Free tier (30 consultations / month) lets individual clinicians test before procurement; residents and interns get unlimited free use; Nabla Connect "embed in days" framing is the lowest-friction Foundation-stage deployment path in the cohort.
Pilot (40-59): yes - Denver Health 50-clinician 8-week pilot is the canonical pilot-stage shape; athenahealth marketplace channel with 30+ provider groups is a low-friction pilot path that does not require Epic-deep procurement.
Scale (60-79): yes - Strong fit. 85K+ clinicians across 130-150+ health organizations + Denver Health 400-clinician post-pilot rollout + UC San Diego Health + LCMC + Iowa + Carle Health demonstrate Scale-stage operational capacity. Slightly behind Abridge at very-large-IDN tier; ahead of Suki / Ambience / DeepScribe on documented named-customer scale.
Optimization (80+): yes - Agentic / Nabla Connect / AMI Labs partnership / multi-language / privacy-architecture differentiation is Optimization-shaped. Post-Series-C roadmap (real-time coding agent, context-aware historical-data agent, nursing / inpatient expansion) is precisely the Optimization feature set.
One-line verdict: Cohort's only peer-reviewed RCT-validated ambient scribe that beat Microsoft DAX Copilot head-to-head; combines the broadest language coverage (35+), the strongest privacy architecture (default-no-audio-storage + 14-day note retention), and the cohort's most aggressive pricing band - at the cost of Epic-native depth, HITRUST disclosure, and a published LLM stack.

Headline numbers

Metric	Value	Evidence
Peer-reviewed RCT (Nabla-specific, head-to-head)	UCLA NEJM AI Nov 2025: 238 physicians, 14 specialties, 72K encounters; Nabla -9.5% time-in-note vs control (95% CI: -17.2% to -1.8%, P=0.02); DAX Copilot -1.7% (P=0.66); Nabla 7.8% faster than DAX (P=0.05)	[THIRD-PARTY - https://ai.nejm.org/doi/abs/10.1056/AIoa2501000, https://pmc.ncbi.nlm.nih.gov/articles/PMC12768499/]
Health-organization footprint	130-150+ healthcare organizations (mid-2025 to early-2026)	[VENDOR-CLAIMED - https://www.nabla.com/, https://www.nabla.com/blog/70m-series-c]
Clinician footprint	85,000+ clinicians	[VENDOR-CLAIMED - https://www.nabla.com/]
Annual encounter volume	20 million+ patient encounters per year	[VENDOR-CLAIMED - https://www.nabla.com/]
Revenue growth	5x in the 6 months prior to Series C (June 2025)	[VENDOR-CLAIMED - https://news.crunchbase.com/ai/nabla-ai-copilot-agents-healthcare-funding/]
Language coverage	35+ languages including bilingual encounters - broadest in cohort	[VENDOR-CLAIMED - https://www.nabla.com/athenahealth, https://glass.health/compare/nabla]
Specialty coverage	55+ specialties; UCLA RCT validated across 14 specialties	[VENDOR-CLAIMED - https://www.nabla.com/athenahealth]
Named customers (US health systems)	Denver Health (400+ clinicians post-pilot), Carle Health, University of Iowa Health Care, LCMC Health, Aultman Health, M Health Fairview, UC San Diego Health, MUSC Children's Health, Children's Hospital Los Angeles, CVS Health	[VENDOR-CLAIMED - https://www.nabla.com/, https://www.nabla.com/press, https://www.nabla.com/case-studies/denver-health]
OEM / white-label channel	Nabla Connect (Oct 2025) powers NextGen Healthcare's "Ambient Assist" and Greenway Health's ambient docs as white-label - only productized OEM-embedding channel in cohort	[VENDOR-CLAIMED + THIRD-PARTY - https://s10.ai/blog/best-ai-medical-scribe-integration-ehr]
Denver Health outcomes (peer-reviewed-grade case study)	40% reduction in per-encounter note-typing time; 30% burnout reduction sustained at 30 and 90 days; 13% pajama-time reduction; 15-point patient satisfaction lift; 300K+ post-rollout encounters	[VENDOR-CLAIMED - https://www.nabla.com/case-studies/denver-health]
Carle Health outcomes	55% of clinicians saved 1+ hour daily on documentation	[VENDOR-CLAIMED - https://www.healthcareaiguy.com/p/company-deep-dive-nabla]
University of Iowa Health Care outcomes	30% burnout reduction sustained at 30-90 days	[VENDOR-CLAIMED - https://www.healthcareaiguy.com/p/company-deep-dive-nabla]
athenahealth marketplace adoption	30+ provider groups via athenahealth Marketplace in first year (Sep 2024 to Oct 2025)	[VENDOR-CLAIMED - https://www.prnewswire.com/news-releases/30-provider-groups-adopt-nablas-ambient-ai-through-the-athenahealth-marketplace-in-first-year-302579950.html]
Free tier?	Yes - 30 consultations / month (active clinicians); unlimited for residents / interns. Cohort-rare.	[THIRD-PARTY - https://glass.health/compare/nabla]
Pricing band (per provider per month, enterprise)	$150-$350 (EHR Source 2026 mid-market band) - 30-60% below Abridge's $600-$800	[THIRD-PARTY - https://www.ehrsource.com/articles/ambient-ai-scribes-comparison/]
Pricing detail (self-serve tiers)	Starter ~$119 / month; Pro ~$239 / month	[THIRD-PARTY - https://glass.health/compare/nabla]
Underlying model architecture	"Proprietary language models and speech-to-text - not a GPT wrapper" per vendor positioning; Azure speech-to-text confirmed as sub-processor; full LLM stack at task-level granularity undisclosed	[VENDOR-CLAIMED + THIRD-PARTY - https://glass.health/compare/nabla, https://trust.nabla.com/]
Privacy architecture (cohort-leading)	Default no audio storage; transcripts only briefly in clinician browser; clinical notes retained 14 days by default, configurable by customer	[VENDOR-CLAIMED - https://trust.nabla.com/]
EU posture (cohort-rare)	GDPR + EU AI Act certifications listed; Belgium data residency option for non-US clients	[VENDOR-CLAIMED - https://trust.nabla.com/]

Dimension scores

Dimension	Score	Weight	Weighted	Evidence
Clinical accuracy + safety	3/4	20	15.0	[THIRD-PARTY + VENDOR-CLAIMED] UCLA NEJM AI RCT (Nov 2025) is the only Nabla-specific peer-reviewed RCT in the cohort: 238 physicians, 14 specialties, 72K encounters, statistically significant 9.5% time-in-note reduction vs control (P=0.02), 7.8% faster than DAX Copilot (P=0.05). Standard clinician sign-off model; vendor-stated 95-98% speech-recognition accuracy band; one Grade 1 mild adverse event in the RCT (extensive patient counseling omitted from A/P + patient instructions). Inaccuracy ratings on 5-point scale: Nabla 2.8 vs DAX 2.7 (P=0.68, not different). Gaps blocking 4/4: no published MedQA / MedHELM / MedBench scores, no FDA SaMD clearance, no Nabla-internal confabulation-detection whitepaper analogous to Abridge's, no word-level "Linked Evidence"-style provenance UI, single-institution RCT (multi-site validation in progress is unknown). Sits at anchor 3 ("at least one peer-reviewed validation study, citations surfaced in product output, hard-gated clinician sign-off") - the UCLA RCT clears the anchor cleanly; anchor 4 blocked by absence of regulatory clearance + multi-site replication. (nabla.md §"Clinical accuracy + safety") - https://ai.nejm.org/doi/abs/10.1056/AIoa2501000, https://pmc.ncbi.nlm.nih.gov/articles/PMC12768499/, https://www.uclahealth.org/news/release/ucla-study-finds-ai-scribes-may-reduce-documentation-time
EHR integration depth	3/4	15	11.25	[VENDOR-CLAIMED + THIRD-PARTY] 20+ named EHRs; native athenahealth Marketplace listing with one-click chart write (HPI / PE / A/P / patient instructions); Oracle Cerner integration named-customer production at Aultman Health (Jan 2026); Epic integration customer-deployed at Denver Health + Carle Health; NextGen + Greenway as white-label OEM via Nabla Connect; Arya + Altera also cited. Nabla Connect (Oct 2025) is the only productized OEM-embedding API in cohort. SMART on FHIR login support cited on athenahealth integration page. Gaps blocking 4/4: Epic Showroom / App Orchard / Pal-tier status unconfirmed (Abridge holds first-Pal designation; Ambience holds Epic Toolbox + Hyperdrive embed); no public FHIR conformance statement; FHIR resources written not enumerated publicly; HIE / TEFCA QHIN participation not surfaced; RCM clearinghouse connector list not enumerated. Sits at anchor 3 ("App Orchard listing on at least one major EHR with bi-directional write on notes + orders + problem list") - clears via athenahealth Marketplace + Cerner deployment + Epic customer deployments; anchor 4 ("production integrations across multiple major EHRs with bi-directional write across clinical and orders resources") blocked by Epic-tier-status gap + FHIR-resource enumeration gap. (nabla.md §"EHR integration depth") - https://athenaconnect.athenahealth.com/marketplace/product/nabla, https://www.nabla.com/athenahealth, https://www.nabla.com/press, https://s10.ai/blog/best-ai-medical-scribe-integration-ehr
Workflow fit + clinician burden reduction	4/4	15	15.0	[THIRD-PARTY + VENDOR-CLAIMED] UCLA NEJM AI RCT controlled-design evidence (the strongest in cohort): -9.5% time-in-note vs control, P=0.02; cohort burden-reduction secondary outcomes (Mini-Z 2.0 +2.69, PFI work exhaustion -0.23) directionally positive for any-scribe arm. Denver Health: 40% per-encounter time reduction, 30% sustained burnout reduction at 90 days, 13% pajama-time reduction, 300K+ post-rollout encounters across 12 specialties (physicians + nurses + APPs + dentists + community health workers + social workers + residents + PTs). Iowa: 30% burnout reduction at 30-90 days. Carle Health: 55% of clinicians save 1+ hour daily. knownwell: 24% per-visit documentation time reduction. Cross-customer: 55% of users save 1+ hour daily; 1.5x more appointments monthly; 27% burnout reduction. 55+ specialties; 35+ languages (broadest in cohort); Denver Health spans 12 specialties including non-physician roles. Sits at anchor 4 ("broad specialty + multi-discipline coverage with controlled-design peer-reviewed validation + customer-published burden-reduction figures with named institutions and methodology") - clears via UCLA RCT + Denver Health multi-discipline + Iowa peer-validated + 35+ language breadth; this is the strongest workflow-fit evidence base in the cohort. (nabla.md §"Workflow fit + clinician burden reduction") - https://ai.nejm.org/doi/abs/10.1056/AIoa2501000, https://www.nabla.com/case-studies/denver-health, https://www.healthcareaiguy.com/p/company-deep-dive-nabla
Compliance + PHI posture	3/4	15	11.25	[VENDOR-CLAIMED + UNKNOWN gaps] Trust center is unusually thorough: SOC 2 Type II (Nov 2024 - Oct 2025 audit period - the most precise SOC 2 date among cohort tear-sheet vendors), ISO/IEC 27001 (Sep 2025), HIPAA + BAA, GDPR, EU AI Act, CCPA, Texas RAMP Level 2 (Jan 2026), PIPEDA, CIS Level 2. Sub-processor list disclosed (GCP US Central + Azure US for US data; GCP Belgium for non-US) - more thorough than any other cohort tear-sheet vendor researched. Default no audio storage + 14-day note retention (configurable) is cohort-leading privacy architecture. Annual third-party penetration testing with critical / high findings tracked to resolution. Gaps blocking 4/4: HITRUST CSF not listed (the rubric's anchor-3 / anchor-4 distinguishing certification); model-training opt-out default on de-identified clinical notes not explicitly enumerated; ISO 42001 (AI management) absent; FedRAMP / FINRA / PCI-DSS not listed; free tier reportedly does not include BAA. Sits at anchor 3 ("BAA + SOC 2 Type II + ISO 27001 + thorough sub-processor disclosure + privacy-by-default architecture") - clears cleanly on architectural privacy + breadth of certifications; anchor 4 blocked by HITRUST + model-training-opt-out gaps. (nabla.md §"Compliance + PHI posture") - https://trust.nabla.com/, https://www.nabla.com/security
Ease of data integration + accuracy (sub-score A: data integration)	3/4	(within 25)	(within 18.75)	[VENDOR-CLAIMED + THIRD-PARTY + UNKNOWN gaps] Native athenahealth Marketplace + Oracle Cerner production deployment + Epic customer deployments + NextGen Ambient Assist + Greenway as white-label OEM via Nabla Connect (Oct 2025, the only productized OEM-embedding channel in cohort). SMART on FHIR login support cited. 20+ EHR claim across third-party reviews. Gaps: no public FHIR R4 conformance statement, no public SDKs analogous to a developer platform, no public-facing developer sandbox, no HIE / TEFCA QHIN participation surfaced, RCM clearinghouse connectors not enumerated, Epic Showroom / Pal-tier status unconfirmed. Anchor 3 ("public FHIR support, native integrations to 1 major EHR + 2-3 RCM / clearinghouse platforms or equivalent EHR-channel breadth") - clears on athenahealth + Cerner + Epic + NextGen + Greenway breadth; falls short of anchor 4 on missing FHIR conformance + clearinghouse enumeration + Epic-Pal-tier confirmation. (nabla.md §"EHR integration depth")
Ease of data integration + accuracy (sub-score B: output + model accuracy)	3/4	(within 25)	(within 18.75)	[THIRD-PARTY + VENDOR-CLAIMED + UNKNOWN] UCLA NEJM AI RCT is the strongest task-effectiveness measurement in cohort - statistically significant 9.5% time-in-note reduction (P=0.02), 7.8% faster than DAX Copilot (P=0.05). Vendor-stated "95-98% speech-recognition accuracy" band across third-party comparisons. UCLA RCT inaccuracy rating 2.8/5 ("occasionally") vs DAX 2.7/5 (P=0.68, not statistically different). Vendor positioning: "proprietary language models and speech-to-text - not a GPT wrapper"; Yann LeCun AMI Labs partnership for world-model architecture. Azure speech-to-text confirmed in sub-processor list. Gaps blocking 4/4: underlying LLM stack at task-level granularity not publicly disclosed (frontier-provider orchestration vs fully proprietary inference is the open question); no published MedQA / MedHELM / MedBench / MedConceptsQA scores; no Nabla-internal confabulation-detection whitepaper analogous to Abridge's. Anchor 3 ("strong second-tier or fine-tuned medical model with cited accuracy + peer-reviewed clinical-effectiveness measurement") - clears strongly via the UCLA RCT; anchor 4 ("frontier-class LLM with named-customer accuracy benchmarks on the buyer's specific clinical task plus competitive MedQA / MedHELM performance") blocked by absence of standardized-benchmark publication + LLM-stack disclosure gap. (nabla.md §"Clinical accuracy + safety") - https://pmc.ncbi.nlm.nih.gov/articles/PMC12768499/, https://glass.health/compare/nabla
Ease of data integration + accuracy (final = average of A + B)	3/4	25	18.75	Average of sub-scores A (3/4) and B (3/4) per rubric instruction.
Cost economics	3/4	5	3.75	[THIRD-PARTY + VENDOR-CLAIMED] Most-disclosed pricing in cohort. EHR Source 2026 enterprise band $150-$350 / provider / month - 30-60% below Abridge's $600-$800. Self-serve tiers published with cross-source confirmation: Free (30 consultations / mo), Starter ($119 / mo), Pro ($239 / mo). Free tier for residents / interns is cohort-rare. Captured-ROI evidence is unusually strong: Denver Health 40% per-encounter time reduction across 300K+ encounters; Carle Health 55% of clinicians save 1+ hour daily; Iowa 30% burnout reduction at 90 days; knownwell 24% per-visit time reduction. Implied per-clinician ROI multiple is well above 5x at every band of pricing even when softened against the UCLA RCT effect size (41 seconds per note vs 18 in control). Triangulation math (5x revenue growth pre-Series-C on ~85K clinicians) is consistent with the $150-$350 band rather than the higher Abridge end. Gap blocking 4/4: no published enterprise rate card (custom-quote), no published implementation-fee schedule. Sits at anchor 3 ("multiple deployment-size scenarios documented; per-tier published; transparent self-serve pricing") - clears cleanly on self-serve transparency + cross-source confirmation; anchor 4 ("fully published rate card including implementation, professional services, and OEM channel terms") blocked by enterprise-quote opacity. (nabla.md §"Pricing") - https://www.ehrsource.com/articles/ambient-ai-scribes-comparison/, https://glass.health/compare/nabla, https://www.nabla.com/case-studies/denver-health
Time-to-value	3/4	5	3.75	[VENDOR-CLAIMED + THIRD-PARTY] Free tier signup gives clinician-level first-note within minutes; athenahealth Marketplace install path measured in days. Denver Health 50-clinician 8-week pilot expanded to 400 clinicians within one week post-pilot - the fastest scale-out timeline documented in cohort case-study literature. UCLA RCT physicians were live within the 60-day intervention window. Nabla Connect's "embed in days" framing for OEM partners is differentiated. Iowa burnout reductions sustained at 30 and 90 days - implying meaningful use within the first month. knownwell 64% increase in clinician adoption over 6 months. Gap blocking 4/4: enterprise-rollout-to-first-clinician-live SLA from contract signature not published; full IDN-scale rollout (LCMC, M Health Fairview, UC San Diego) measured in quarters not weeks at largest scale. Sits at anchor 3 ("under 12 weeks to meaningful use at typical pilot size; multi-channel self-serve paths") - clears via Denver Health 8-week pilot + athenahealth marketplace path + free tier; anchor 4 ("under 8 weeks to first clinician live at enterprise rollout scale") plausible at single-clinic pilot inside existing customer; not yet established at enterprise rollout scale. (nabla.md §"Workflow fit + clinician burden reduction") - https://www.nabla.com/case-studies/denver-health, https://pmc.ncbi.nlm.nih.gov/articles/PMC12768499/
Total		100	78.75

Pricing detail

Source: Cross-referenced from third-party scribe-comparison aggregators (glass.health, EHR Source 2026 comparison, Commure, marvix.ai, DeepScribe 2026 ambient comparison). Nabla does not publish a public list-price page; the per-tier figures below are cross-checked across 3+ independent sources. All [THIRD-PARTY estimated] unless tagged.

Free tier: $0 / month - 30 consultations / month (active clinicians); unlimited for residents / interns with verified institutional email. No HIPAA BAA on this tier. Cohort-rare - only Suki has any comparable self-serve tier; Nabla's is the broader. - https://glass.health/compare/nabla
Starter (Solo): ~$119 / clinician / month - ambient note generation, basic EHR export (athenahealth marketplace + browser), 35+ languages, custom templates, HIPAA BAA. - https://glass.health/compare/nabla
Pro: ~$239 / clinician / month - specialty templates, real-time ICD-10 / HCC / MCC coding suggestions, SSO / SAML, multi-EHR integration depth, dictation support. - https://glass.health/compare/nabla
Enterprise: Custom pricing - mid-market scribe-comparison band $150-$350 / provider / month per EHR Source 2026 enterprise comparison. Cohort-comparable range against Abridge ($600-$800), DAX Copilot ($369-$1,512 depending on Microsoft reseller channel), Suki ($299-$399). Nabla anchors at the cohort's most aggressive enterprise pricing - roughly 30-60% below Abridge at typical list. - https://www.ehrsource.com/articles/ambient-ai-scribes-comparison/
Implementation / integration fees: [UNKNOWN] No public list. Implied to be lower than the Epic-Inside class because the architecture is browser-first; Nabla Connect OEM channel sold as "embed in days." Worth raising in right-of-reply.
Triangulation math: 5x revenue growth in the 6 months before the Series C close (June 2025) on a Series-B-era 20K-provider footprint that grew to 85K+ clinicians by Q2 2025 (4.25x clinician-base expansion). Implied per-clinician annualized revenue consistent with the $150-$350 per-month enterprise band rather than the higher Abridge end. - https://www.nabla.com/blog/70m-series-c
OEM / white-label channel pricing: [UNKNOWN] Nabla Connect terms not publicly enumerated. NextGen Ambient Assist and Greenway ambient docs OEM partnerships have undisclosed revenue-share / per-seat-fee structures.

Integrations

athenahealth (cohort-leading marketplace channel): [VENDOR-CLAIMED + THIRD-PARTY] Native Marketplace listing (Sep 2024), 30+ provider groups via this channel in first year, one-click chart write across HPI / PE / A/P / patient instructions, SMART on FHIR login support. Named customers via this channel: knownwell, Pomelo Care, Homeward Health, MUSC Children's Health, Wayspring, Mankato Clinic. - https://athenaconnect.athenahealth.com/marketplace/product/nabla, https://www.prnewswire.com/news-releases/30-provider-groups-adopt-nablas-ambient-ai-through-the-athenahealth-marketplace-in-first-year-302579950.html
Oracle Health / Cerner: [VENDOR-CLAIMED] Aultman Health System January 2026 named-customer production deployment via Oracle Cerner integration - first explicitly-named Cerner-deep production rollout in cohort outside Microsoft DAX. CODE program participation tier unconfirmed. - https://www.nabla.com/press
Epic: [VENDOR-CLAIMED + THIRD-PARTY] Customer-deployed at Denver Health + Carle Health; integration is real and named-customer-production. [UNKNOWN] Epic Showroom / App Orchard / Pal-tier formal status unconfirmed in our research window. Worth raising in right-of-reply.
NextGen Healthcare (white-label OEM): [VENDOR-CLAIMED + THIRD-PARTY] Nabla powers NextGen Healthcare's "Ambient Assist" product as white-label OEM - the deepest OEM-channel relationship in cohort. NextGen end-customers receive Nabla's product behind a different brand. - https://s10.ai/blog/best-ai-medical-scribe-integration-ehr
Greenway Health (white-label OEM): [VENDOR-CLAIMED + THIRD-PARTY] Nabla powers Greenway Health's ambient documentation as white-label OEM.
Nabla Connect (OEM embedding API): [VENDOR-CLAIMED + THIRD-PARTY] Launched October 2025 - generic embedding API for "any EHR vendor to embed Nabla in days." Only productized OEM-embedding channel in cohort. - https://s10.ai/blog/best-ai-medical-scribe-integration-ehr
Other named EHRs: [VENDOR-CLAIMED] Arya, Altera cited; 20+ EHRs claimed in aggregate.
SMART on FHIR + SSO/SAML/SCIM: [VENDOR-CLAIMED] Cited on athenahealth integration page. No public FHIR conformance statement enumerating versions / resources / launch contexts.
FHIR R4 conformance + specific resources written: [UNKNOWN] Not enumerated publicly. Worth raising in right-of-reply.
HIE / TEFCA QHIN participation: [UNKNOWN] Not surfaced.
RCM / clearinghouse connectors: [VENDOR-CLAIMED + UNKNOWN depth] Real-time ICD-10 / HCC / MCC coding-suggestion product is in market via athenahealth integration. Specific clearinghouse / payor connectors (Waystar / Change Healthcare / Optum / Availity / Epic Resolute / Cerner Revenue Cycle) not enumerated on a public integrations page.
Developer surface: Nabla Connect provides documented embedding API for OEM partners; closer to partnership-channel API than public-developer-portal SDK.
CDS content partnerships: [UNKNOWN] No explicit UpToDate / NEJM / JAMA content-into-product partnerships analogous to Abridge's 2026 announcements.

Editorial assessment

Nabla is the cohort's most scientifically-credentialed mid-market ambient-AI scribe and the only vendor in the category whose product is named in a published, peer-reviewed, head-to-head randomized controlled trial against a major competitor in NEJM AI. The November 2025 UCLA RCT - 238 physicians, 14 specialties, 72,000 encounters - measured a statistically significant 9.5% reduction in time-in-note for Nabla vs control (P=0.02) and a 7.8% faster reduction than Microsoft DAX Copilot (P=0.05). That is the single strongest piece of independent clinical evidence in the cohort. Abridge's strongest peer-reviewed paper has no control arm; the strongest controlled-design multi-vendor study (Mass General Brigham + UCSF, 5 hospitals) does not separate by product; Ambience's KLAS-validated revenue-uplift is customer-side observational. Nabla has the cohort's only head-to-head RCT, and Nabla won it.

The structural differentiation around that evidence is also real. Nabla ships 35+ languages (the broadest in cohort), a default-no-audio-storage privacy architecture that competitors cannot match without re-architecting, the only productized OEM-embedding API (Nabla Connect powers NextGen Ambient Assist and Greenway Health as white-label), an unusually thorough public sub-processor disclosure (GCP US Central + Azure US for US; GCP Belgium for non-US), and EU-rooted GDPR + EU AI Act certifications that a Paris-founded vendor naturally inherits but US-founded competitors have to build toward. The pricing band ($150-$350 per provider per month per third-party scribe comparisons; per-tier $119 / $239 list at Starter / Pro) is 30-60% below Abridge and Ambience at typical enterprise list, and Nabla is the only cohort vendor with both a self-serve free tier (for residents and clinicians under 30-consultation cap) and a productized OEM channel.

Three structural gaps sit underneath that strength. First, the underlying model stack is not publicly disclosed at task-level granularity. Azure speech-to-text in the sub-processor list confirms at least the STT layer is Azure-hosted; whether the production LLM inference is fully proprietary (as the "not a GPT wrapper" vendor positioning claims) or orchestrates over frontier providers (OpenAI / Anthropic / Google) for portions of the workflow is the open question. There is no MedQA / MedHELM / MedBench publication that would let an academically-minded buyer benchmark Nabla against the cohort on standardized accuracy. The UCLA RCT is a clinical-effectiveness measurement, not a standardized accuracy benchmark. Second, the compliance posture has visible gaps despite being unusually thorough overall - HITRUST CSF is not surfaced on the public trust center, and the model-training opt-out default on de-identified clinical notes is undocumented. The default-no-audio-storage architecture is the cohort's strongest architectural privacy defense; the contractual-default question is the remaining open one. Third, Epic Showroom / App Orchard / Pal-tier status is unconfirmed. Epic integration is real and customer-deployed (Denver Health, Carle Health), but the formal Epic-partnership tier is not surfaced - and for buyers anchored on Epic-Pal status as a procurement signal, that is the gap that pushes them toward Abridge.

The cost economics are the best-disclosed in the cohort. EHR Source's 2026 enterprise-comparison band of $150-$350 per provider per month puts Nabla at roughly 30-60% of Abridge's list and meaningfully below Ambience and DAX Copilot. The per-tier Starter / Pro pricing of $119 / $239 is published with enough cross-source confirmation to be trustworthy. The captured-ROI evidence is unusually strong: Denver Health documents 40% per-encounter time reduction across a 300,000+ encounter post-rollout sample; Carle Health documents 55% of clinicians saving 1+ hour daily; Iowa documents 30% burnout reduction sustained at 90 days. Even softening the vendor-claim aggregates against the controlled UCLA RCT effect size (41 seconds per note), the per-clinician ROI multiple is well above 5x at every band of pricing.

When to revisit: when Nabla publishes (a) Epic Showroom / App Orchard / Pal-tier status, (b) HITRUST CSF tier and audit date, (c) explicit underlying LLM stack disclosure at task-level granularity, (d) documented model-training opt-out default on de-identified notes, (e) a second independent peer-reviewed RCT (different institution, EHR, specialty mix), (f) MedHELM / MedQA / MedBench performance, (g) FHIR R4 conformance statement with enumerated resources. Any one closes a current gap; (a) and (b) together would push the case for Nabla as cohort-overall co-leader with Abridge rather than the credentialed mid-market alternative.

Best for

Stage: Foundation, Pilot, Scale, and Optimization - cleanest end-to-end stage fit in the cohort. Foundation-stage fit is strongest of any cohort peer (free tier + Nabla Connect "embed in days"). Slightly behind Abridge at the very-large-IDN Scale tier (Kaiser 24K class) but ahead at every other stage.
Company profile: Mid-size and regional health systems with multi-EHR delivery networks; safety-net systems with multilingual patient populations; OEM partners building ambient documentation into their own products; academic medical centers that value peer-reviewed RCT evidence and price discipline more than KLAS ranking and named-peer reference; behavioral health buyers in two-party-consent states that need a default-no-audio-storage architecture.
Industry fit: Strong for hospital / health-system (cohort primary, especially mid-size and regional), ambulatory groups (cohort-leading via athenahealth Marketplace channel), FQHC / community health / safety-net (cohort-leading via 35+ language coverage + browser-only architecture, Denver Health proof point), behavioral health (default-no-audio-storage architecture + Wayspring named customer), pediatrics (MUSC Children's Health + CHLA named), nursing (Denver Health multi-discipline pilot), multi-national / cross-Atlantic operations (only cohort vendor with native EU + Belgium residency posture). Moderate for ASC / surgical, Emergency Medicine. Weak for imaging / radiology (out of product surface).
Sales motion: Hybrid - free tier + self-serve Starter/Pro for individual clinicians and small practices; athenahealth Marketplace path for ambulatory groups; enterprise procurement for health-system rollouts. The OEM / white-label channel via Nabla Connect is a unique route that did not exist in cohort before October 2025.
Annual tooling budget: Highly variable by tier and channel. Self-serve Starter at $119 / clinician / month = ~$1,400 / year per clinician. Enterprise band at $150-$350 / clinician / month = ~$1,800-$4,200 / year per clinician (30-60% discount vs Abridge at the same scale). At 250-clinician footprint: ~$450K-$1.05M / year mid-band. At 2,500-clinician footprint: ~$4.5M-$10.5M / year mid-band.
Skip if: You need (a) Epic-Pal-tier formal designation as a procurement signal (Abridge holds that, Nabla's Epic-tier status is unconfirmed), (b) KLAS Best-in-KLAS / KLAS-validated ROI documentation as a procurement signal (Abridge holds #1 KLAS 2025+2026; Ambience holds KLAS Trailblazer 2026; Nabla's KLAS standing is not surfaced), (c) HITRUST r2 as a procurement gate (not on public trust center), (d) word-level Linked-Evidence-style provenance UI (Abridge's), (e) FDA-cleared CDS depth (Nabla positions CDS as roadmap requiring FDA approval, not current), (f) Kaiser-scale 24K-clinician deployment precedent at the largest IDN tier (Abridge holds that), (g) a single-vendor RCM coding + clearinghouse + payor stack (Iodine / Waystar territory, not Nabla's primary surface).

Right-of-reply

Nabla will receive this tear-sheet seven calendar days before publication of the Yardstick Research 2026 Yardstick Report, including all measured numbers, sample outputs, and editorial assessment. Nabla will be given the opportunity to flag factual errors - incorrect pricing, misquoted feature availability, outdated screenshots, factual misstatement in the editorial assessment. Nabla will not be given the opportunity to request a score revision, dispute the rubric or its weights, withdraw from inclusion, negotiate ranking placement, or suggest changes to the editorial assessment beyond factual correction. Where Nabla flags a factual correction, the correction will be applied if verified and noted here; where Nabla disputes scoring, the dispute will be recorded in the appendix but the score will stand. Silence from the vendor during the right-of-reply window will be treated as no objection.

Right-of-reply gaps

Specific [UNKNOWN] items surfaced in the dossier and explicitly raised with the vendor in right-of-reply:

Epic Showroom / App Orchard / Pal-tier status. Epic integration is customer-deployed (Denver Health, Carle Health) but the formal Epic-partnership tier is not surfaced publicly. Listed at any tier? At Pal-tier depth comparable to Abridge?
Underlying LLM stack at task-level granularity. Is the production inference fully proprietary, or does it orchestrate over frontier providers (OpenAI / Anthropic / Google) for portions of the workflow? If frontier-provider, which models and on which workloads, and where do those API calls land geographically?
HITRUST CSF certification. Held or not held? If held, which tier (e1 / i1 / r2)? Audit date? Given the unusually thorough certification list on the trust center, omission feels like a real gap rather than a disclosure gap.
Model-training opt-out default on de-identified clinical notes. Audio default is no-storage. What is the default for de-identified note text? Is opt-out configurable in the standard MSA / DPA?
MedHELM / MedQA / MedBench / MedConceptsQA performance. Are there published standardized-benchmark numbers Nabla can share?
FDA SaMD positioning. With the AMI Labs / agentic-autonomy roadmap and real-time coding capabilities, is Nabla moving toward regulated-device claims? Current regulatory positioning?
FHIR R4 conformance + specific resources written. Which FHIR resources does Nabla write back (DocumentReference / MedicationRequest / ServiceRequest / Encounter / Observation)? Is there a published conformance statement?
Oracle Health CODE program participation. Aultman Health is Cerner-integrated; what is the CODE program tier?
RCM / clearinghouse connector list. Waystar, Change Healthcare / Optum, Availity, Epic Resolute, Cerner Revenue Cycle - which are live native connectors?
HIE / TEFCA QHIN participation. Live?
Two-party-consent recording-law overlay (CIPA, FL, PA). What is the contractual handling of patient-consent capture between Nabla and customer, beyond the structural default-no-audio-storage defense?
Current headcount. Tracxn's 143-144 (February 2026) is the most recent third-party number; current number?
HQ legal-entity confirmation. Nabla Technologies, Inc. for US operations (Brooklyn) - relationship to the French operating entity (Paris)? Single corporate parent or sister entities?
Pricing transparency. Willingness to publish a per-tier enterprise rate-card range, or does enterprise pricing remain custom-quote only?
Implementation-fee schedule. Public list?
ISO 42001 (AI management). Held? Pursuing?
Second independent peer-reviewed RCT. The UCLA NEJM AI trial is single-institution; is a multi-site or different-institution validation in progress?
KLAS Arch Collaborative / Best-in-KLAS Ambient Speech standing. Does Nabla participate in KLAS evaluations? Current standing?
Named CISO / Chief Compliance Officer. Not on the public team page; who owns information security and HIPAA compliance internally?
The 2020 GPT-3 medical-Q&A research. Is the company comfortable publishing a current-day reflection on the research, the lessons it informed for the ambient-documentation product, and any guardrails it shaped? This would convert a residual brand-image risk into a credibility asset.

Sources

Nabla first-party: - https://www.nabla.com/ - https://www.nabla.com/team - https://www.nabla.com/press - https://www.nabla.com/blog - https://www.nabla.com/blog/70m-series-c - https://www.nabla.com/security - https://www.nabla.com/athenahealth - https://www.nabla.com/case-studies - https://www.nabla.com/case-studies/denver-health - https://www.nabla.com/case-studies/carle-health - https://www.nabla.com/press-release/nabla-raises-70m-series-c-to-deliver-agentic-ai-to-the-heart-of-clinical-workflows-bringing-total-funding-to-120m - https://trust.nabla.com/ (SafeBase-hosted) - https://help.nabla.com/en/articles/4854594

Press releases / investor / news: - https://www.prnewswire.com/news-releases/nabla-raises-70m-series-c-to-deliver-agentic-ai-to-the-heart-of-clinical-workflows-bringing-total-funding-to-120m-302483646.html - https://www.prnewswire.com/news-releases/30-provider-groups-adopt-nablas-ambient-ai-through-the-athenahealth-marketplace-in-first-year-302579950.html - https://www.prnewswire.com/news-releases/nabla-raises-24m-in-series-b-to-fuel-expansion-of-its-ambient-ai-assistant-to-transform-care-delivery-302027132.html - https://www.prnewswire.com/news-releases/ending-clinician-burnout-global-summit-2025-welcomes-nabla-as-title-sponsor-to-drive-systemic-change-302559873.html - https://www.statnews.com/2025/06/17/nabla-raises-70-million-ambient-market-heats-up/ - https://news.crunchbase.com/ai/nabla-ai-copilot-agents-healthcare-funding/ - https://siliconangle.com/2025/06/17/clinical-ai-startup-nabla-focus-agentic-automation-raising-70m-funding/ - https://www.builtinnyc.com/articles/nabla-secures-70m-new-funding-20250618 - https://hospitalogy.com/articles/2025-06-18/breaking-down-nabla-sword-and-commons-clinics-recent-raises/ - https://hospitalogy.com/articles/2025-02-12/nabla-restoring-the-joy-of-medicine/ - https://hlth.com/insights/news/nabla-raises-70m-series-c-to-develop-agentic-ai-assistant-for-healthcare-2025-06-18 - https://www.highlandeurope.com/nabla-raises-70m-series-c-to-deliver-agentic-ai-to-the-heart-of-clinical-workflows-bringing-total-funding-to-120m/ - https://www.linkedin.com/pulse/nabla-raises-70m-series-c-advance-ambient-scribe-margaretta-colangelo-tfrlf - https://www.fiercehealthcare.com/ai-and-machine-learning/nabla-banks-70m-series-c - https://www.fiercehealthcare.com/ai-and-machine-learning/carle-health-teams-nabla-ai-scribe-assistant - https://www.fiercehealthcare.com/health-tech/navina-and-nabla-unveil-partnership-integrate-clinical-copilot-ambient-ai - https://techcrunch.com/tag/nabla/

Third-party analyst profiles: - https://www.crunchbase.com/organization/nabla - https://www.cbinsights.com/company/nabla - https://www.cbinsights.com/company/nabla/financials - https://pitchbook.com/profiles/company/265249-00 - https://www.privco.com/company/Nabla-Technologies - https://www.zoominfo.com/c/nabla/397709641 - https://leadiq.com/c/nabla/5c47255d1d0000cf9f8190b8 - https://tracxn.com/d/companies/nabla/__9HH3-1wWpzD4sdQ2lByXePi71KyKD3MkB_6kBcu2pQw - https://startupintros.com/orgs/nabla - https://canvasbusinessmodel.com/blogs/owners/nabla-who-owns - https://www.healthcareaiguy.com/p/company-deep-dive-nabla

Third-party reviews / customer-side comparisons: - https://www.ehrsource.com/articles/ambient-ai-scribes-comparison/ - https://glass.health/compare/nabla - https://glass.health/resources/best-ai-medical-scribe - https://www.marvix.ai/blog/11-best-ai-scribe-tools-in-2026 - https://www.commure.com/blog-scribe/ai-medical-scribe-alternatives - https://patientnotes.ai/resources/best-ai-scribes - https://orbdoc.com/compare/ai-medical-scribe-comparison-2025/ - https://bastiongpt.com/which-is-the-best-ai-scribe - https://digitalhealth.folio3.com/blog/best-ai-medical-scribe-top-picks/ - https://getscribe.commure.com/blog/best-ai-medical-scribes-2026-top-10-tools-compared - https://www.deepscribe.ai/resources/best-ai-medical-scribes-for-athenahealth-2026 - https://www.deepscribe.ai/resources/best-ai-medical-scribes - https://scribeable.ai/compare/nabla - https://www.plaud.ai/blogs/articles/ai-note-taker-doctors-software-vs-hardware - https://www.soapnoteai.com/soap-note-guides-and-example/ambient-ai-scribe-adoption-2026/ - https://s10.ai/blog/best-ai-medical-scribe-integration-ehr

Peer-reviewed studies + research literature: - https://ai.nejm.org/doi/abs/10.1056/AIoa2501000 (UCLA NEJM AI RCT, Nov 26 2025 - the canonical Nabla-specific RCT) - https://pmc.ncbi.nlm.nih.gov/articles/PMC12768499/ (PMC mirror of UCLA NEJM AI RCT) - https://pubmed.ncbi.nlm.nih.gov/40672471/ (PubMed indexing of UCLA RCT) - https://www.medrxiv.org/content/10.1101/2025.07.10.25331333v1 (medRxiv preprint of UCLA RCT, July 2025) - https://pmc.ncbi.nlm.nih.gov/articles/PMC12265753/ (PMC mirror of medRxiv preprint) - https://www.researchgate.net/publication/393627906_A_Randomized-Clinical_Trial_of_Two_Ambient_Artificial_Intelligence_Scribes_Measuring_Documentation_Efficiency_and_Physician_Burnout (ResearchGate listing) - https://www.uclahealth.org/news/release/ucla-study-finds-ai-scribes-may-reduce-documentation-time (UCLA Health press release) - https://www.pulmonologyadvisor.com/news/ai-based-ambient-scribes-may-reduce-physician-documentation-time-burnout/ - https://www.ophthalmologyadvisor.com/news/ai-based-ambient-scribes-may-reduce-physician-documentation-time-burnout/ - https://distilinfo.com/2026/04/20/ai-scribes-cut-clinical-documentation-time-significantly/ - https://www.llmsfordoctors.com/trials/ambient-ai-scribes-dax-nabla-rct-2025/ - https://www.emarketer.com/content/ai-scribes-raise-roi-questions-after-minimal-time-saved-study - https://digitalhealthwire.com/ai-scribes-arent-productivity-tools-yet/ - https://www.nature.com/articles/s41746-025-01670-7 (Nature / npj Digital Medicine framework on clinical safety + hallucination rates - cohort-level calibration)

Integration + EHR marketplace: - https://athenaconnect.athenahealth.com/marketplace/product/nabla (athenahealth Marketplace listing) - https://amia.org/membership/partners/nabla (AMIA partner listing)

Historic 2020 GPT-3 medical-Q&A research artifact: - https://www.theregister.com/2020/10/28/gpt3_medical_chatbot_experiment/ - https://www.artificialintelligence-news.com/news/medical-chatbot-openai-gpt3-patient-kill-themselves/ - https://news.ycombinator.com/item?id=26280727

Trust + compliance: - https://www.keragon.com/hipaa/hipaa-compliant-checker/nabla - https://www.nabla.com/blog/all-you-need-to-know-about-nablas-privacy-and-security-features - https://www.nabla.com/legal-notice - https://www.nabla.com/legal-documents